Wall Street’s trash contains buried treasure

In a paper that starts out by stating that “no one enjoys getting dumped," two investing quants reveal some surprising, and potentially lucrative, traits of companies that have really let themselves go. With about half of the money invested in American stocks now sitting in index funds, and many active managers holding portfolios that resemble them—just try beating the market these days without “Magnificent 7" stocks such as Nvidia or Microsoft—index castoffs have a hard time meeting someone new.

That is when investors should pounce, says Rob Arnott, chairman of advisory firm Research Affiliates, with colleague Forrest Henslee. This week they are unveiling a stock index named NIXT that would have earned investors about 74 times their money since 1991 by buying stocks kicked out of indexes.

The basic idea isn’t revolutionary: Plenty of hedge funds make money, or try to, by buying stocks that are about to enter a widely held index such as the S&P 500 and selling short those about to leave. They know that there will be forced buyers and sellers who don’t care about the price. Instead of a speculative one-night stand with those stocks, though, NIXT includes them after they have been dumped. And, while not being wedded to those losers forever, it dates them for five years—an eternity for fast-money types. The surprising thing is how long the good performance lasts.

A helpful way to see the effect is to think of a nixed stock’s value relative to the S&P 500 as being equal to one on the day it gets booted. Former members of the S&P 500, Russell 1000 and Nasdaq-100 indexes were worth 2.25 as much, on average, a year before deletion—quite a fall. But then they start to do better than the index and keep it up for five years, reaching 1.28. In other words, they beat the market by a chunky 5% a year.

Companies that get added to an index tend to rally beforehand and then lag behind it for the first year or so, but not by nearly as much. Why the difference? Those that get dumped are also more thinly traded. Then a bunch of funds have to sell them all at once.

“Part of what you’re seeing is a bounce from that pressure," says Arnott. “It’s quite a marvelous pattern."

Arnott embraces index funds, which tend to beat most active managers over the years, but he is a luminary in the world of finance nerds constantly trying to build a better mousetrap. A paper he wrote in 2005 with Jason Hsu and Philip Moore helped ignite the smart beta trend that now guides more than one trillion dollars in assets. Research Affiliates, founded just three years earlier, says it has created strategies now used by funds with about $150 billion under management.

The basic idea of Arnott’s original paper was to create an index based on fundamental factors such as profitability instead of market value. Its main U.S. stock index beat a standard capitalization-weighted one by 1.8 percentage points a year from 1992 through 2022—a substantial edge. The one tracking small companies was even better, edging the comparable index by 2.7 points a year.

Among other advantages, fundamental indexes are more likely to hold on to nixed stocks. Investors in plain vanilla index funds and happy to stick with them shouldn’t be feeling FOMO, though: The effect of missing out on nixed companies is tiny because they are literally a rounding error when they get dumped. But, for those looking for something spicier, nixed ones on their own might be especially attractive now.

Stocks that get dumped are by their nature small and cheap. Between 1991 and 2022, deletions traded at a 26% discount to the S&P 500 in terms of their price-to-earnings ratio while additions fetched an 83% premium. And since Arnott predicts that small-cap value stocks are poised to outperform the S&P 500 after lagging behind it for years, he thinks the new index also could be well-positioned.

What’s the catch? Unlike a fundamental index, which can absorb plenty of money, there are going to be limits to how much it can track nixed stocks before the effect diminishes.

“We’ve modeled it out to $1 billion, and it works fine," says Arnott.

Another catch: There is a cheap, widely followed index out there that has done even better. Investors who bought and held the Nasdaq-100, made up mostly of large U.S. tech stocks, have endured some wild bubbles and wrenching drawdowns but made about three times as much as the S&P 500 between 1991 and 2023.

Investors who believe that artificial intelligence is just getting started can keep concentrating their bets on indexes stuffed with glamour stocks. But those willing to take a chance on short, dark and ugly strangers might be pleasantly surprised.

Write to Spencer Jakab at Spencer.Jakab@wsj.com